Strength in Weakness: The Mutable Collagenous Tissue of Echinoderms

Abstract

Mutable collagenous tissue (MCT) is a type of connective tissue that is characterized by its capacity to undergo rapid, nervously mediated changes in mechanical properties. In terms of both the magnitude of these changes and the timescale within which they occur (less than one second to a few minutes), this tissue appears to be unique to the phylum Echinodermata and, as it is ubiquitous in all five extant echinoderm classes, it represents one of the four major defining features of the phylum, together with pentaradial symmetry, endoskeletal stereom (calcite meshwork), and the water vascular system. MCT has been the subject of intensive scientific investigation for over 50 years. The primary aim of this contribution is to provide a comprehensive and definitive survey of the current state of knowledge of this remarkable tissue. After outlining the history of the scientific investigation of MCT, we review current information on its anatomical distribution, organization at the histological, ultrastructural and molecular levels, and physiology—focusing on its mechanical behavior and the regulation of this behavior; its significance for echinoderm biology, including pathology; and biomedical and other applications that exploit MCT-derived components or biological principles. We conclude by drawing attention to more serious deficiencies in the current knowledge base and suggesting how these should be rectified.

1. Introduction

It has long been recognized that the collagens constitute the most abundant and widely distributed protein family in the animal kingdom [1]. According to a classification based on domain structure and supramolecular organization, 28 distinct collagen proteins have been identified, the most prevalent of which occur as parallel fibril-like aggregates or three-dimensional meshworks in the extracellular matrix (ECM) of connective tissue [2,3]. Connective tissue consisting mainly of fibril- or meshwork-forming collagens is present in all but the simplest multicellular animals. It is by far the most important structural material in the bodies of all non-arthropods [4,5], has essential regulatory functions by virtue of its involvement in cell adhesion, differentiation, and intercellular signaling [6,7,8], and in some animals may function as a body-wide mechanosensitive signaling network [9].

As a structural material, the roles of collagenous connective tissue are mainly to resist, transmit, and dissipate mechanical forces and to store and release elastic strain energy [10]. However, the specific tensile properties of different collagenous structures vary according to the functional demands for which these structures are adapted [3,11,12,13]. The tensile properties of individual structures may also change during development and aging and during homeostatic and other physiological processes. Such changes usually occur within timescales of hours to years, are due primarily to the modulation of the composition and supramolecular organization of the ECM, and are controlled proximally by endocrine, paracrine, and gene regulatory mechanisms [13,14,15,16]. By way of contrast, collagenous structures in echinoderms (starfish, sea-urchins, and their close relatives) can undergo drastic changes in mechanical properties within timescales of less than one second to a few minutes, which result mainly from the neural regulation of factors responsible for cohesion between their collagen fibrils [13,17].

The term ‘mutable’ was first applied to echinoderm collagenous tissue evincing such variable tensility by J.P. Eylers in 1979 (in litt.). It is also known as ‘catch connective tissue’ [18,19], a designation that the present authors avoid because it invites comparison with the ‘catch muscle’ of bivalve molluscs [20] and encourages misconceptions about the possible involvement of actin-myosin-based active contractility in the mechanical adaptability of mutable collagenous structures. It is also inappropriate for certain such structures that exhibit only irreversible destabilization and are likely to lack ‘catch’-like activity, i.e., reversible stiffening and destiffening (see Section 3).

Mutable collagenous tissue (MCT) occurs widely throughout the five extant echinoderm classes, i.e., Asteroidea (starfish), Crinoidea (sea-lilies and featherstars), Echinoidea (sea-urchins), Holothuroidea (sea-cucumbers), and Ophiuroidea (brittlestars and basket stars) [17], and there is indirect evidence for its presence in extinct Paleozoic echinoderm classes [21,22]. Since it also appears to be unique to the Echinodermata, it represents one of the major defining features of the phylum, together with pentaradial symmetry, endoskeletal stereom (calcite meshwork), and the water vascular system [23]. MCT plays an important role in many facets of echinoderm biology including locomotion, defense, alimentation, and reproduction, and is conjectured to have been a major contributor to the evolutionary success of the phylum [24,25,26].

This review first outlines the history of the scientific investigation of MCT, then summarizes current knowledge of its anatomical distribution, organization at the histological, ultrastructural, and molecular levels and physiology—focusing on its mechanical behavior and the regulation of this behavior. It continues with a consideration of the pervasive significance of MCT for the biology of echinoderms in health and disease, and an overview of biomedical and other applications that exploit MCT-derived components or biological principles. It concludes by drawing attention to the more serious deficiencies in current knowledge of MCT biology and suggesting how these should be rectified. The review cites 140 papers on MCT biology, around 60% of the total published since interest in this subject was reawakened in the 1960s (see Section 2). Parts of Section 4 and Section 5 are adapted and updated from refs. [13,17].

2. History of MCT Research

The history of the scientific investigation of MCT began over 150 years ago when advances in optical microscopy and staining technology allowed the histological organization of animals to be examined systematically [27]. This revealed an apparent mismatch between the morphology and the functioning of certain echinoderm anatomical components. Jickeli [28] proposed that in featherstars the active motility of the appendages known as cirri and the downward power-stroke of their swimming arms could result only from the action of structures that had the histological appearance and staining properties of connective tissue. Perrier [29] called these structures ‘hyaline muscles’ and provided histological evidence for their innervation, which was questioned by Bosshard [30] but confirmed by Reichensperger [31], who also attributed the autotomy (defensive self-detachment) of featherstar cirri and arms to the ‘hypercontraction’ of such tissues. Lindemann [32] investigated the physicochemical mechanism underlying the changes in consistency—hardening, softening, and liquefaction—exhibited by the dermis of certain sea-cucumbers. He described this phenomenon as being ‘… a process sui generis which is very probably unique in the whole animal kingdom’ (translated). In vitro experiments on strips of sea-cucumber body wall enabled Jordan [33,34] to demonstrate nervous influences on its mechanical behavior. He concluded that it was ‘… midway between connective and muscular tissue’ (translated), although connective tissue alone was revealed by his histological investigations. Von Uexküll [35] regarded sea-cucumber dermis unequivocally as a catch musculature (‘Sperrmuskulatur’). Von Uexküll [36] also had no doubts that, since it could immobilize the spine-test joint, the spine ligament of sea-urchins was a form of catch musculature, thus disagreeing with the view of other authors that it was connective tissue [37,38]. These soft tissue structures in featherstars, sea-cucumbers, and sea-urchins thus demonstrated a discrepancy between morphological data, which indicated they were conventional connective tissue, and physiological data, which suggested the presence of contractile cells.

After the 1920s, this conundrum was neglected for 40 years, until publications by Serra-von Buddenbrock on sea-cucumber dermis [39], Takahashi on the sea-urchin spine ligament [40,41], and Meyer on featherstar ligaments [42] stimulated a revival of interest that led to a steady expansion in knowledge of various aspects of MCT and the eventual recognition of its importance for the biology of all echinoderms [43].

Whilst research activity on MCT over the subsequent 50 years has concentrated on its basic biology, with a persistent and continuing focus on the molecular mechanism(s) modulating its tensility [43,44,45] and on the involvement of specialized neurons—juxtaligamental cells—as effectors of tensile change [46,47], consideration has also been given to the potential biomedical and engineering applications of MCT. There has been growing interest in MCT as a source of components for the synthesis of new materials intended for therapeutic interventions and as inspiration for the design of entirely artificial materials and devices with adjustable mechanical properties [27,48,49,50].

3. Anatomical Distribution of MCT

Table 1. Mutable collagenous structures. The list is limited to those structures in which the capacity for variable tensility has been demonstrated experimentally or confirmed by direct observation. Tensile changes: Id, irreversible destiffening associated with autotomy; Id*, irreversible destiffening associated with opportunistic self-detachment (see Section 6.2); Is, irreversible stiffening; R, reversible stiffening and destiffening. APL, arm plate ligament; IM, intervertebral muscle; LI, longitudinal interambulacral; PRM-LBWM, pharyngeal retractor muscle–longitudinal body wall muscle; PTG, peripheral through-going; SP, spine-pedicel.

Class and Species	Anatomical Component	Mutable Collagenous Structure	Tensile Change	Reference
Asteroidea
Acanthaster planci	Spine joint	Dermis and SP ligament	R	[52]
Asterias rubens	Body wall	Dorsolateral dermis	R,Id	[53]
Pycnopodia helianthoides	Body wall	LI ligament	R,Id	[54]
Marthasterias glacialis	Tube foot	Dermis	R	[55]
Crinoidea
Antedon bifida	Cirrus	Synathrial ligament	R	[56]
Antedon mediterranea	Cirrus	Synostosal ligament	Id	[57]
Cenocrinus asterius	Stalk	Symplexal/PTG ligaments	R	[58]
Cenocrinus asterius	Stalk	Synostosal ligament	Id	[58]
Antedon mediterranea	Arm	Diarthrial ligament	R	[59]
Antedon mediterranea	Arm	Syzygial ligament	Id	[59]
Himerometra robustipinna	Visceral mass	Tegmen and septa	Id	[60]
Echinoidea
Anthocidaris crassispina	Spine joint	Capsular ligament	R	[61]
Diadema antillarum	Spine joint	Capsular ligament	Id*	[62]
Diadema setosum	Spine joint	Central ligament	R	[63]
Diadema setosum	Aristotle’s lantern	Tooth ligament	R	[64]
Paracentrotus lividus	Aristotle’s lantern	Compass depressor ligament	R	[65]
Paracentrotus lividus	Tube foot	Dermis	R	[55]
Paracentrotus lividus	Body wall	Peristomial membrane	R	[66]
Echinus esculentus	Body wall	Periproctal dermis	R	[67]
Lytechinus variegatus	Globiferous pedicellaria	Head-stalk ligament	Id	[68]
Holothuroidea
Stichopus chloronotus	Main body wall	Dermis	R	[69]
Stichopus chloronotus	Main body wall	Dermis	Id*	[25]
Eupentacta quinquesemita	Introvert body wall	Dermis	Id	[70]
Eupentacta quinquesemita	LBWM	PRM-LBWM tendon	Id	[70]
Holothuria forskali	Cuvierian tubule	Inner connective tissue layer	Is	[71]
Ophiuroidea
Ophiocomella ophiactoides	Dorsal disk body wall	Dermis	Id	[72]
Ophiomastix lütkeni	Arm	Intervertebral ligament	R	[73]
Ophiocomina nigra	Arm	Intervertebral ligament	R,Id	[74]
Ophiocomina nigra	Arm	IM autotomy tendon	Id	[75]
Ophiura ophiura	Arm	Distal ventral APL	R,Id	[76]
Ophiura ophiura	Arm	Proximal ventral APL	R	[76]
Ophiophragmus filograneus	Arm	Genital plate-lateral APL	Id	[77]
Ophiocomina nigra	Arm spine joint	Spine ligament	R,Id*	[78]
Ophioderma longicaudum	Ventral disk body wall	Oral shield-oral plate ligament	R	[79]

lists those echinoderm collagenous structures in which the capacity for variable tensility has been either observed directly or detected experimentally. An example of the former is afforded by the disintegration of brittlestar intervertebral ligaments during arm autotomy [51]. The experimental demonstration of variable tensility has usually involved the in vitro mechanical testing of isolated tissue preparations, for example before and after their treatment with agents that act on the nervous system, such as elevated [K⁺], anesthetics, and neurotransmitter agonists and antagonists (see Section 5).

Table 2. Non-mutable collagenous structures. The list comprises structures that failed to evince the capacity for variable tensility in physiological experiments.

Class and Species	Anatomical Component	Collagenous Structure	Reference
Echinoidea
Echinus esculentus	Aristotle’s lantern	Compass-rotular ligament	[80]
Eucidaris tribuloides	Spine joint	Central ligament	[81]
Ophiuroidea
Ophiocomina nigra	Arm	Intervertebral muscle non-autotomy tendon	[75]
Ophioderma longicaudum	Interbrachial frame	Radial and interradial oral plate ligaments	[79]

lists the small number of echinoderm collagenous structures that have not yielded evidence for variable tensility when tested in physiological experiments.

Table 1 and Table 2 demonstrate that the vast majority of echinoderm connective tissue structures that have been investigated have been found to show mechanical adaptability. These mutable collagenous structures represent a range of morphofunctional types, including (in all five living classes) the body wall dermis and a variety of ligaments, and (in brittlestars alone) tendons. Whilst it is obvious that a wide diversity of echinoderm collagenous structures has been found to be mutable, data on individual structures have been obtained from, in most cases, a very small number of species. In view of this taxonomic limitation, and the possibility that investigated species are not representative, it is unsafe to make assumptions about the pervasiveness of the mutability phenomenon. A cautionary example is provided by the central spine-joint ligament of sea-urchins, with that of Diadema setosum (belonging to Subclass Euechinoidea) being mutable, but that of Eucidaris tribuloides (belonging to Subclass Cidaroidea) lacking mutability [63,81].

4. Organization of MCT

4.1. Extracellular Components

MCT consists predominantly of extracellular materials and contains a relatively small volume fraction of cellular constituents. In histological sections viewed in the light microscope, the extracellular components of most mutable collagenous structures are seen as aggregations of fibers of varying diameter, conformation, and spatial arrangement (Figure 1A,B and Figure 2A,B). Some, e.g., interossicular ligaments, are predominantly simple parallel fiber arrays, although even these may display structural variability, particularly at insertion regions [65,82]. Others consist of stacks of parallel fiber sheets with the fiber orientation varying between adjacent sheets: the genital bar-lateral arm plate ligament of amphiurid brittlestars is a simple example of this [77], whereas the dermis of both starfish dorsolateral body wall and sea-urchin peristomial membrane has three interconnected but structurally distinct layers dominated by orthogonal fiber arrays [83,84]. Another arrangement is seen in the MCT of sea-urchin and starfish tube feet, which consists of an inner sheath of longitudinal fibers and an outer sheath of crossed-fiber helical arrays [55]. MCT thus demonstrates a microstructural diversity comparable to that of vertebrate fibrous connective tissue [85].

The extracellular fibers of MCT are, with one exception, parallel bundles of transversely banded collagen fibrils accompanied by interfibrillar proteoglycans and loose arrangements of microfibrils (Figure 1C,D and Figure 2C,D). The exception is certain tendon fibers at the autotomy planes of brittlestar arms (see below). The banded collagen fibrils of MCT are parallel assemblages of trimeric collagen molecules arranged in a regular staggered array with adjacent molecules offset by 40 to 80 nm. In most echinoderm fibrils, the collagen molecules consist of two fibrillar α chains (1α and 2α) which form (1α)₂2α heterotrimers [87,88,89]. Some collagen molecules in the fibrils of sea-urchin MCT include a third fibrillar chain (5α) and form (1α)₂5α heterotrimers [87]. The N-propeptide of the 5α chain shows two peculiarities: (1) it is retained after fibril assembly, which is not the case in the echinoderm 2α chain and in all vertebrate fibrillar procollagens [90], and remains at the surface of the fibrils; (2) it contains 11 SURF (‘sea-urchin fibrillar’) modules, which are also present in the 2α N-propeptide and in fibrosurfin—an interfibrillar component of sea-urchin MCT; fibrosurfin is structurally related to the fibropellins that constitute the apical lamina surrounding sea-urchin embryos [91,92].

Analyses of the Strongylocentrotus purpuratus genome and the transcriptome of the body wall of the sea-cucumber Cladolabes schmeltzii have not uncovered any features of the echinoderm ECM that could be correlated with its variable tensility [93,94,95]. These analyses, as well as a targeted proteomic investigation of the body wall of the sea-cucumber Apostichopus japonicus [96] and a comparative genomic analysis of four echinoderm classes [97], have also shown the presence of fibrillin genes and their transcripts or gene products. This complements biochemical and immunological evidence that the microfibrils occurring in MCT and non-mutable echinoderm ligaments [78,80,98] consist at least partly of fibrillin-like proteins [99]. The function of these microfibrils is unclear. They may facilitate slippage between adjacent fibril bundles during MCT deformation, maintain the orientation of fibrils during deformation, and contribute to passive elastic recoil after the removal of external forces.

Other interfibrillar molecules contribute to the linkage between adjacent collagen fibrils and for that reason must have a major influence on interfibrillar stress transfer and the mechanical properties of MCT. Biochemical evidence indicates that proteoglycans, which consist of a protein core and glycosaminoglycan (GAG) sidechains, act as binding sites for other molecules that are inferred to form interfibrillar crossbridges (discussed in Section 5.2). After staining with the cationic dyes cuprolinic blue or cupromeronic blue, which label GAG sidechains, electron-dense deposits are visualized as granules on the fibrillar surface or as interfibrillar bridges, both arranged regularly within each fibril D-period (Figure 3) [86,98,99,100,101]. Fucosylated chondroitin sulfate is the major GAG of sea-cucumber dermis. It has been found only in this tissue, is covalently associated with collagen fibrils, and consists of a chondroitin sulfate core with branches of the sulfated sugar α-L-fucose. The functional significance of the fucose side-branches is not understood, although it is known that their enzyme-inhibitory activity is crucial for the ‘exogenous’ anticoagulant, antithrombotic, and antidiabetic properties of fucosylated chondroitin sulfate [49,101,102].

MCT contains other molecules that contribute to interfibrillar cohesion but whose extracellular disposition is unknown; these are discussed in Section 5.2.

The tendon fibers at the autotomy planes of brittlestar arms are continuous with the basement membrane (BM) of the intervertebral muscles (Figure 4). Their histochemistry and the ultrastructural similarities between their junctions and the muscles and conventional (non-mutable) BM–muscle complexes of echinoderms and vertebrates confirm that they are BM derivatives [13]. Genomic analysis has demonstrated that echinoderms express the full BM extracellular ‘toolkit’—collagen IV, the glycoproteins laminin and nidogen, and the heparan sulfate proteoglycan perlecan [93,95,97]. There is no direct information on the spatial arrangement of these molecules in brittlestar autotomy tendons or conventional echinoderm BMs. Some of the molecules are very similar to their vertebrate homologs, particularly, in the case of collagen IV and nidogen, with respect to domains involved in intermolecular bonding [103,104]. The configuration of echinoderm BMs (including brittlestar autotomy tendons) is therefore likely to resemble that detected in all investigated vertebrates and invertebrate BMs, which consist of separate collagen IV and laminin networks linked to each other and to cell surfaces by molecules that include nidogen and perlecan [105].

4.2. Cellular Components

4.2.1. Juxtaligamental Cells

The most abundant cellular components of MCT are neurosecretory-like cell processes containing large (diameter > 100 nm), membrane-bounded dense-core vesicles (LDCVs) (Figure 1C,E, Figure 2D,E and Figure 4B). Those present in brittlestar intervertebral ligaments were the first to be described and were named ‘juxtaligamental cells’ (JLCs) because their somata are located outside but close to the ligaments [106]. JLCs are the subject of a recent comprehensive review [46] on which the following account is based.

All investigated mutable collagenous structures are permeated by, or adpressed against, cell processes containing LDCVs [46]. In all classes except the Echinoidea there is usually more than one type of process distinguishable mainly by the size and shape of their LDCVs (

Table 3. Ultrastructure of juxtaligamental large dense-core vesicles (LDCVs) associated with two examples of collagenous structures from each class. The placement of LDCV data from a particular species in the columns headed ‘LDCV Type’ is not intended to imply homology with LDCVs of other species in the same column. c, circular; CL, capsular ligament; cr, crescentic; CSL, central spine ligament; CT, connective tissue; IL, intervertebral ligament; JLN, juxtaligamental node; LIL, longitudinal interambulacral ligament; o, oval; TF, tube-foot.

Class and Species	JLC Component	Location	LDCV Type (a)		LDCV Type (b)		Reference
			Shape	Size (nm)	Shape	Size (nm)
Asteroidea
Asterias rubens	Somata	LIL	c-o	≤250 × 380	c	≤540	[86]
	Processes	LIL	c-o	≤230 × 360	c	≤540
Marthasterias glacialis	Somata	TF outer CT sheath		ca. 150		ca. 250	[55] [113]
	Processes	TF outer CT sheath	c-o	ca. 150		ca. 250
Crinoidea
Decametra sp.	Somata	Stereom spaces	c	ca. 100	o	ca. 180	[114]
	Processes	Diarthrial arm ligaments	c	ca. 100	o	ca. 180
Metacrinus rotundus	Somata	Stereom spaces	c	150–200	c	600	[115]
	Processes	Stalk ligaments	c	150–200	c	600
Echinoidea
Arbacia punctulata	Somata	Inside + outside CL	c-o	≤320			[116]
	Processes	Inside + outside CL	c-o	≤220/330
Diadema setosum	Somata	Inside CSL	c-o	100–500			[63]
	Processes	Inside CSL	c-o	100–500
Holothuroidea
Holothuria forskali	Somata	Cuvierian tubule CT	c	200–400	c-o	300–600	[71]
	Processes	Cuvierian tubule CT	c	200–400	c-o	300–600
Stichopus chloronotus	Somata	Body wall dermis	o-cr	≤1400			[107]
	Processes	Body wall dermis	c-o	≤300 × 700
Ophiuroidea
Ophiocomina nigra	Somata	IL JLNs	c	≤160	c-o	≤210 × 460	[74,106]
	Processes	Intervertebral ligament	c	≤300–400	c-o	≤250 × 750
Ophiocomina nigra	Somata	Spine ligament JLN	c	≤150	c-o	≤220 × 310	[78]
	Processes	Spine ligament	c	≤200	c-o	≤270 × 520

; Figure 1E and Figure 2D). The processes occur mostly in compact bundles of two or more, which always comprise more than one type. Although most sea-cucumber juxtaligamental components appear to be separated from the ECM by a basement membrane [107,108,109,110], no such basement membrane is present in the other classes (apart from single starfish and brittlestar exceptions [111,112]).

In all classes, the LDCVs of some juxtaligamental cell bodies and their processes are strongly acidophilic in histological sections (Figure 1B and Figure 2B), a feature that may serve as a marker for juxtaligamental tissue [46]. Ultrastructural evidence from all classes indicates that the LDCVs are produced by the Golgi complex. The consistent presence of microtubules in JLCs and their labeling by antibodies against acetylated tubulin, somatostatin, and synaptotagmin B [78,117,118] suggest that JLCs are neurons. Those with smaller LDCVs in the 100–300 nm size range may be regular aminergic or peptidergic neurons. Those with larger LDCVs > 300 nm in diameter are the main reason why JLCs are often described as being ‘neurosecretory-like’, since LDCVs of a comparable size are often present in the neurosecretory cells of other phyla [119,120,121].

The juxtaligamental somata of all classes other than the Holothuroidea occur in aggregations. Large aggregations—called juxtaligamental nodes—are associated with all brittlestar mutable collagenous structures (Figure 1A,F) and some ligaments of sea-lilies, featherstars and sea-urchins [106,122,123]. Large aggregations are always located outside the MCT into which their juxtaligamental processes project and they have a central neuropil-like region (Figure 1G). The latter intimates that they may have ganglion-like integrative and coordinative functions, which is supported by the presence in brittlestars of a capsule of glia-like cells, which compartmentalize the juxtaligamental somata (Figure 1H) [78]. Starfish MCT contains smaller aggregations of cell bodies that have a micro-ganglion-like organization including a central cluster of cell processes and a loose outer layer of glia-like cells (Figure 2D–F) [86]. Cells sharing features with the radial glial cells of echinoderm radial nerve cords are in close contact with the JLCs of sea-urchin tooth ligaments [124,125]. JLC aggregations appear to be absent from sea-cucumbers. Glia-like cells ensheathing bundles of cell processes have been seen in the MCT of one sea-cucumber [109].

4.2.2. Other Cells

Fibrillar MCT usually includes a sparse population of cell bodies containing heterogeneous vacuoles that resemble lysosomes and enclose cytoplasmic debris or collagen fibrils (Figure 1C) (see, e.g., [65,78,110,126]). They occur also in non-mutable echinoderm connective tissue [127] and, given the dearth of obvious fibroblasts in MCT, at least some may be pluripotential cells that can adopt a fibrogenic phenotype.

Morula cells are abundant in, and peculiar to, sea-cucumber dermis. These contain very large (diameter 3–6 µm) membrane-bounded vesicles containing material of variable electron density, and may synthesize molecules that participate in fibrogenesis [110,128].

Possible regular neuronal processes that lack LDCVs occur in mutable collagenous structures of all five classes [86]. In sea-cucumber dermis, such processes are associated with glia-like cells [110,128].

Myocytes have been observed in several starfish, sea-urchin, and sea-cucumber structures, in most of which they occupy a small proportion of the total cross-sectional area (

Table 4. Mutable collagenous structures that contain myocytes. CSA, proportion of the total cross-sectional area occupied by myocytes; SP, spine-pedicel.

Class and Species	Anatomical Component	Collagenous Structure	CSA (%)	Reference
Asteroidea
Acanthaster planci	Spine joint	Dermis and SP ligament	<2–ca. 30 1	[129]
Echinoidea
Anthocidaris crassispina	Spine joint	Capsular ligament	3	[61]
Eucidaris tribuloides	Spine joint	Capsular ligament	1.5	[81]
Paracentrotus lividus	Aristotle’s lantern	Compass depressor ligament	8	[65]
Stylocidaris affinis	Aristotle’s lantern	Compass depressor ligament	13	[130]
Holothuroidea
Eupentacta quinquesemita	Introvert body wall	Dermis	1–4	[110]

¹ CSA varies from <2% in the dermis that ensheathes the joint to ca. 30% in the ligament between the spine base and the pedicel ossicle of the body wall [129].

).

5. Physiology of MCT

5.1. Passive Mechanical Properties

5.1.1. Baseline Mechanical Behavior

It was established more than 40 years ago [73] that fibrillar MCT can show passive mechanical behavior that generally resembles that of the fibrillar collagenous tissue of vertebrates. Like the latter [131,132], MCT is, at least in certain physiological states, a viscoelastic material that demonstrates tensile creep (elongation) under constant load (Figure 5A–C), stress relaxation (force decay when stretched to and maintained at a fixed length) (Figure 5D), stress–strain curves with non-linear toe regions (Figure 5E,F), and time-dependence—for example, its elastic stiffness (slope of the stress–strain curve) increases with increasing strain rate [61,73,83,133].

Table 5. Mechanical properties of (A) echinoderm mutable collagenous structures and, for comparison, (B) selected mammalian collagenous structures: creep data. Values shown are means or ranges. Coefficient of viscosity is strain rate/stress. CL, cruciate ligament; DW, distilled water; K, elevated [K⁺]; ---, no data.

Class and Species	Mutable Collagenous Structure	Stress 1 MPa	Coefficient of Viscosity MPa·s	Strain at Breakage	Reference
A. Echinodermata
Asteroidea
Asterias rubens	Basal aboral body wall 2	---	0.5	---	[53]
Asterias rubens	Distal aboral body wall 2	---	1.4	---	[53]
Echinoidea
Diadema setosum	Central spine ligament	ca. 0.1	20–6000	0.8–3.7	[63]
Paracentrotus lividus	Compass depressor ligament	0.17–1.16	104–1477	1.41–4.54	[136]
Holothuroidea
Actinopyga echinites	Dermis	0.056	100	---	[137]
Apostichopus japonicus	Dermis	0.3–3.0	0.076–35	---	[138]
Eupentacta quinquesemita	Introvert dermis	0.03	ca. 100	≤9	[70]
Holothuria leucospilota	Dermis	0.056	11	---	[137]
Stichopus chloronotus	Dermis (untreated, K-stiffened)	6 × 10−5	0.06, 1.25	---	[139]
Thyone inermis	Dermis 3 (untreated, DW-stiffened)	---	5100, 15,800	---	[140]
Ophiuroidea
Ophiocomella ophiactoides	Aboral disk body wall 2	---	ca. 1000	---	[72]
Ophiocomina nigra	Intervertebral ligament	0.2–1.1	500–7100	0.6–3.0	[74]
B. Chordata
Mammalia
Homo sapiens	Anterior CL, posterior CL	---	39.29, 48.67	---	[141]
Homo sapiens	Patellar tendon	---	438.1	---	[141]
Mus musculus	Skin 4	---	1375	---	[142]
Rattus norvegicus	Uterine cervix (pregnant)	0.029	128.9	---	[143]

¹ In some of the cited investigations, ‘stress’ is initial stress, i.e., load/initial cross-sectional area; in others, it is constant stress, achieved by continuously adjusting the load to compensate for the progressive decline in cross-sectional area during sample extension. ² The mechanical properties of these structures are determined mainly by the dermis. This, however, consists partly of endoskeletal ossicles, which confounds comparison with purely collagenous structures. ³ The experimental animals were immersed in distilled water for at least 30 min before the excision, treatment (further exposure to distilled water), and mechanical testing of dermis samples, which may explain the anomalously high ‘untreated’ viscosity value. ⁴ The mechanical properties of mammalian skin are determined mainly by the dermis [144].

and

Table 6. Mechanical properties of (A) echinoderm mutable collagenous structures and, for comparison, (B) selected mammalian collagenous structures: stress–strain data and ultimate properties. Values shown are means or ranges. Elastic stiffness is the slope of the linear portion of the stress–strain curve. Ultimate stress (=tensile strength) is the maximum stress observed during a stress–strain test. Ultimate strain is the strain occurring at maximum stress. ACh, acetylcholine; K, elevated [K⁺]; ---, no data.

Class and Species	Mutable Collagenous Structure	Strain Rate s−1	Elastic Stiffness MPa	Ultimate Stress MPa	Ultimate Strain	Reference
A. Echinodermata
Asteroidea
Asterias rubens	Aboral body wall 1 (anesthetized)	0.03	2.48	0.68	0.49	[145]
Echinaster spinulosus	Aboral body wall 1 (stiffened by handling)	0.01–0.02	249–353	37–45	12–15	[83]
Coscinasterias calamaria	Aboral body wall 1 (anesthetized)	0.09–0.24	8.6–10.6	3.27	0.43	[146]
Linckia laevigata	Dermis 1	0.0004, 0.004	20.9, 36.0	3.65, 6.27	30.5, 23.0	[147]
Crinoidea
Metacrinus rotundus	Aboral brachial ligament	---	0.4–15.9	---	---	[148]
Echinoidea
Anthocidaris crassispina	Capsular spine ligament (ACh-stiffened)	ca. 0.003–0.29	230–420	18–38	ca. 0.3	[61]
Eucidaris tribuloides	Capsular spine ligament	0.006	25	8	---	[149]
Eucidaris tribuloides	Capsular spine ligament (stiff state)	---	200	---	---	[150]
Paracentrotus lividus	Compass depressor ligament	0.003–0.250	3.3–44.2	1.5–23.2	0.56–6.50	[136]
Holothuroidea
Actinopyga echinites	Dermis (untreated, K-stiffened)	1.5	1.67, 2.08	---	---	[138]
Holothuria forskali	Cuvierian tubule (quiescent, elongated)	---	0.044, 19.7	0.88, 3.2	2.0, 0.27	[71]
Holothuria leucospilota	Dermis (untreated, K-stiffened)	1.5	0.42, 0.59	---	---	[138]
Stichopus chloronotus	Dermis (untreated, K-stiffened)	1.5	0.024, 0.128	0.03, 0.115	>12–13, ---	[139]
Ophiuroidea
Ophiocomina nigra	Intervertebral ligament (anesthetized)	---	---	6.17	---	[74]
B. Chordata
Mammalia
Bos taurus	Digital extensor tendon	0.1	639	95.7	0.231	[151]
Homo sapiens	Dermis (cryopreserved)	0.1, 10	15–41, 7–45	---	---	[152]
Homo sapiens	Spine anterior longitudinal ligament	0.5	50	32	1.15	[153]
Rattus norvegicus	Tail tendon	0.0013	1304	40–80	0.05–0.17	[154]

¹ See Table 5, Footer ².

provide data on the creep and stress–strain behavior of mutable collagenous structures and a selection of mammalian collagenous structures. With respect to all the featured mechanical properties (viscosity, elastic stiffness, etc.), there is considerable quantitative variation between different mutable collagenous structures, some of which may be attributable to differences in ECM composition and microarchitecture: dermis consists of a 3-D fiber meshwork, whereas ligaments are parallel fiber arrays. There is also considerable quantitative variation between homologous mutable collagenous structures in different species, which in some cases may reflect the varying functional demands for which they are adapted (compare data for Actinopyga echinites and Holothuria leucospilota in Table 5, and Anthocidaris crassispina and Paracentrotus lividus in Table 6). In comparison with mammalian collagenous tissue, MCT exhibits a much wider range of values for viscosity, generally lower values for elastic stiffness and ultimate stress but much higher values for ultimate strain.

5.1.2. Tensile Change

The capacity for tensile change varies qualitatively between different mutable collagenous structures. Each shows one of four patterns of tensile change (Table 1): (1) only irreversible destabilization (as occurs during autotomy), (2) irreversible destabilization, as well as reversible stiffening and destiffening, (3) only reversible stiffening and destiffening, and (4) only irreversible stiffening [44,155].

The dramatic and irreversible loss of tensile strength occurring at autotomy can be mimicked experimentally and quantified using isolated tissue preparations undergoing creep tests (Table 5). Treatment with neuro-active agents such as elevated [K⁺] causes an abrupt decrease in viscosity and tissue rupture (Figure 5A).

Reversible changes in viscosity have been demonstrated in creep tests and stress relaxation tests (Table 6; Figure 5B,C). Reversible changes in tensile strength, tensile stiffness and other parameters have been investigated by means of standard stress–strain tests (Figure 5E,F) and by dynamic testing methods in which samples are subjected to oscillating strain (Figure 5G–I). The so far unique example of irreversible stiffening—exhibited by sea-cucumber Cuvierian organs—has been quantified in stress–strain tests (Figure 6). Data from biomechanical studies have been used to generate a number of models that aim to specify the contribution of different extracellular components to mechanical behavior and reveal their role, if any, in variable tensility. The currently prevalent model, derived originally from work on sea-cucumber dermis, is based on the concept that the tissue can adopt three mechanical states—standard, stiff and soft (Figure 5G–I). These have qualitatively different features: only dermis in the soft state shows stress-induced softening; dermis in the soft and standard states is characterized by J-shaped stress–strain curves with prominent toe regions, while the stress–strain curve of stiff dermis is linear [19,73,135].

5.2. Molecular Basis of Tensile Change

Variation in the capacity of MCT for tensile change is probably enabled by an important difference between MCT and most collagenous tissues of vertebrates: cohesion between adjacent collagen fibrils in the latter is highly dependent on permanent, covalent molecular interactions and is relatively stable [156,157], in contrast to MCT, in which interfibrillar cohesion is less dependent on covalent interactions and is labile. This explains the observations that echinoderm collagen fibrils can be extracted from whole tissue by mild chemical and mechanical methods [158,159], whereas the isolation of intact collagen fibrils from normal adult vertebrate collagenous tissues requires harsher treatments such as enzymolysis and gross mechanical manipulation [160,161,162]. A major aim of MCT research has been the elucidation of the determinants of interfibrillar cohesion and how these determinants are modulated.

During the past 50 years, several hypotheses for the molecular basis of variable tensility have been proposed and subsequently superseded (as recounted in previous reviews [25,73,81,99,163,164,165,166]). Current ideas on possible molecular mechanisms are derived primarily from investigations of reversible changes in sea-cucumber dermis, assume the premises of the three-state paradigm mentioned above (see Section 5.1.2), and focus on chemical factors that can be isolated from the dermis and that influence its mechanical behavior in vitro (Figure 7), though their specific roles in vivo have not been determined with certainty. The tensilins are the most thoroughly investigated of these factors. Tensilins have a high degree of sequence identity to TIMP (tissue inhibitor of metalloproteinase) proteins and can be isolated from dermis only by procedures that cause cell lysis, which suggests they are present in intracellular stores. They stiffen samples of whole dermis and aggregate isolated collagen fibrils in vitro [108]. A comparison of the properties of intact and truncated recombinant tensilin suggested that the N-terminal TIMP-like domain interacts strongly with GAGs attached to the surface of collagen fibrils, and that interfibrillar linkages may involve the dimerization or oligomerization of tensilin mediated by its C-terminal regions [44] (Figure 8).

In vitro experiments have indicated that sea-cucumber tensilins have only a partial stiffening effect in that they convert dermis from the soft to the standard state but not from the standard to the stiff state [167]. The phylum-wide importance of tensilin is also questionable, in view of the inconsistent effect of sea-urchin tensilin-like protein on the mutable compass depressor ligament of the sea-urchin Paracentrotus lividus [136]. In contrast to the tensilins, a second stiffening protein—‘novel stiffening factor’ (NSF)—switches sea-cucumber dermis from the standard to the stiff state, has no effect on soft dermis, and does not aggregate isolated collagen fibrils [168]. The different actions of tensilin and NSF are compatible with the idea that separate molecular mechanisms are responsible for the soft-to-standard and standard-to-stiff transitions. Another protein—stiparin—aggregates collagen fibrils, but, unlike the tensilins, does not affect the whole dermis. It can be extracted by prolonged immersion of dermis in seawater alone and may be a constitutive component that contributes to a basal level of interfibrillar cohesion [169].

Three destiffening molecules have been extracted from sea-cucumber dermis and partially characterized: ‘stiparin inhibitor’, a 62 kDa sulfated glycoprotein that binds stiparin and inhibits its fibril aggregating activity; ‘plasticiser’, a small (<15 kDa) protein that directly destiffens dermal ECM and is released only after cytolytic treatments [169]; and softenin, a ca. 20 kDa protein that can be extracted without cell lysis. Softenin may compete for tensilin binding sites on the collagen fibrils, since it disaggregates tensilin-aggregated collagen fibrils in vitro and reversibly destiffens cell-dead dermis that is in the standard state [170].

Other possible MCT destiffeners include matrix metalloproteinases (MMPs), which have important roles in ECM remodeling associated with echinoderm development and regeneration [171,172,173]. The synthetic MMP inhibitor galardin stiffens sea-urchin compass depressor ligaments in all three mechanical states, the effect being much weaker on stiff ligaments than on soft and standard ligaments. One explanation for these findings is that ligament stiffness is modified through crosslink degradation by constitutive MMP activity regulated by the cellular release of an endogenous MMP inhibitor [174].

As discussed below in Section 5.3.3, several endogenous peptides have indirect, cell-mediated effects on the mechanical behavior of MCT preparations. However, it has been speculated that holokinin-1 and holokinin-2, whose heptapeptide sequences are present in the C-terminal domain of 5α collagen (see Section 4.1), may destiffen the dermis by directly disrupting as yet unidentified interfibrillar linkages involving 5α collagen [175].

Further evidence that separate mechanisms produce the soft-to-standard and standard-to-stiff changes has been acquired by comparing the water content and distribution in MCT samples that are in different mechanical states. In the case of sea-cucumber dermis and sea-urchin compass depressor ligaments, it has been demonstrated that water moves out of the tissues during the standard-to-stiff shift but not during the soft-to-standard shift [176,177], and that the standard-to-stiff transition alone is accompanied by a significant increase in the collagen fibril packing density [98,178]. In both these transitions in sea-cucumber dermis there is also a significant increase in the number of crossbridges connecting adjacent collagen fibrils, as revealed by transmission electron microscopy [178]. On the basis of the latter observations, Tamori et al. [178] proposed that dermal stiffening is achieved by three mechanisms: (1) increased crossbridge formation in both transitions; (2) tensilin-dependent stiffening of collagen fibrils (through subfibril fusion) in the soft-to-standard transition; and (3) in the standard-to-stiff transition, increased bonding within the proteoglycan matrix between fibril bundles, which increases fibril packing density and squeezes out water.

The model of Tamori et al. [178] incorporates the hypothesis that the mechanical adaptability of sea-cucumber dermis is achieved partly by changes in the stiffness of collagen fibrils consequent on their disaggregation into subfibrils and subsequent reaggregation. A similar mechanism was previously proposed on the basis of ultrastructural evidence for fibril disaggregation in featherstar ligaments [100]; however, this has not been reported for any other mutable collagenous structure. More recently, in a series of investigations using time-resolved synchrotron small-angle X-ray scattering (SAXS) to compare the behavior of sea-cucumber dermis in three mechanical states, Gupta and co-workers moved from an initial model in which the variable tensility of the tissue was attributed to modulation of interfibrillar matrix stiffness alone [179,180,181] to one that also incorporates changes in fibril stiffness mediated by a chemoelastic process involving the reversible scission of collagen chains (at unspecified chemical bonds) [45]. It is thus apparent that, despite sea-cucumber dermis being the most intensively investigated mutable collagenous structure, the molecular mechanisms responsible for its variable tensility have yet to be established with certainty. At a more general level, the phylum-wide applicability of information and concepts derived from research on sea-cucumber dermis is at present obscured by the scarcity of relevant information on the MCT of other echinoderm classes [43].

5.3. Regulation of Mechanical Properties

5.3.1. Juxtaligamental Cells

Circumstantial evidence suggests that JLCs are the effectors that directly alter the tensile properties of MCT: JLC processes extend into and terminate within MCT, have no feasible cellular targets in the tissue, and constitute a pathway between the ECM and the motor nervous system, and they have not been found in non-mutable echinoderm collagenous structures [75,80,81]. In addition, possible effector molecules occur in sea-cucumber JLCs—the stiffening proteins tensilin and stiparin—have been immunolocalized to LDCVs in the dermis [169] and tensilin has been detected by immunohistochemistry and in situ hybridization in JLCs of Cuvierian tubule connective tissue [71]. Crucially, it has yet to be demonstrated that dermal stiffening is preceded or accompanied by the release of tensilin, or any other putative effector agent, from JLCs. A correlation between alterations of tensile state and changes in JLC ultrastructure has been observed in a minority of mutable collagenous structures. Such changes have been associated mainly with irreversible destabilization or stiffening and usually include evidence that LDCVs or their contents are released into the extracellular compartment [60,71,98,182].

LDCV-containing processes tend to occur as clusters in which usually two types of process are arranged in parallel and in close proximity (see Section 4.2.1), possibly indicating a functional relationship [78,113,147,182]. Whilst this might be a reciprocal effector system, with one cell-type destabilizing the ECM and the other type stabilizing it, the possibility that processes containing LDCVs in the smaller size range (see Table 3) are aminergic or peptidergic neurons suggests that their role may be to regulate the effector activity of neurosecretory-like JLCs. However, the LDCVs of vertebrate neurosecretory neurons are known to handle a variety of molecules involved in a wide range of biological processes, including ECM degradation [183,184,185]. It is therefore feasible that each of the different cell-types in MCT could have both effector and regulatory functions.

5.3.2. Nervous System

There is a considerable amount of morphological and physiological information on the neural regulation of MCT tensility. As discussed above (see Section 4.2.1), the JLCs, which are associated with all mutable collagenous structures, are likely to be neurons. In all classes there is morphological evidence for a functional linkage between juxtaligamental components and the central nervous system, which at the histological level is represented by the extension into juxtaligamental tissue of neural pathways from the central nervous system [46,186,187]. The best characterized of these are the complex supply of motor nerves from the radial nerve cord to brittlestar juxtaligamental nodes [106,118,182] and the simpler pathways serving sea-lily, featherstar, and sea-urchin JLC aggregations [122,123]. Ultrastructural investigations have revealed possible synaptic junctions between JLCs and exogenous neuronal processes in sea-lilies, featherstars, sea-urchins, and brittlestars [78,122,188,189]. There is variation between the echinoderm classes regarding which neural subsystems provide motor pathways to JLCs: the hyponeural subsystem is involved in starfish and brittlestars [106,187]; the aboral (apical) subsystem in sea-lilies and featherstars [122]; the ectoneural subsystem in sea-cucumbers [186,190]; and the ectoneural and hyponeural subsystems in sea-urchins [191].

The neuropharmacology of MCT has attracted considerable interest. Acetylcholine (ACh) is the main excitatory transmitter in echinoderm motor systems [192]. In vitro pharmacological experiments have shown different patterns of cholinergic responses. ACh and nicotinic receptor agonists cause an initial increase in viscosity and later decrease in the viscosity of sea-cucumber dermis, whereas muscarinic agonists cause a decrease in viscosity [193]. These findings suggest that both stiffening and destiffening of sea-cucumber dermis are controlled by cholinergic motoneurons. Sea-urchin capsular spine ligaments are stiffened by ACh (Figure 5E) and by both nicotinic and muscarinic agonists, the latter having a slower effect [194]. In contrast, both types of cholinomimetic agonists destiffen the cirral ligaments of a sea-lily, the muscarinic effect again being slower [133]. The body wall of a starfish is stiffened by ACh and muscarinic agonists, but not by nicotinic agonists [53].

Several other neurotransmitters have been shown to affect the mechanical behavior of isolated MCT preparations in vitro. The catecholamines adrenaline, noradrenaline, and dopamine cause a biphasic (destiffening then stiffening) response in sea-urchin spine ligaments, as does β-phenylethylamine, although its derivatives, tyramine and octopamine, cause only destiffening [194]. Glutamate-like immunoreactivity has been demonstrated in neuron-like cells and within mutable ligaments in the arms of a featherstar. L-glutamate causes irreversible destiffening of MCT-containing ligaments at syzygies (interossicular junctions specialized for autotomy) in the arms of this animal, which provides evidence that the neural control of autotomy in featherstars involves a glutamatergic pathway. ACh depresses the response of syzygial preparations to L-glutamate [59].

It is to be expected that the nervous system both controls the mechanical properties of MCT and enables changes in these properties to be coordinated with muscle activity. Such coordination has been demonstrated using the sea-urchin spine joint as a model system [195]. Each spine joint is encircled by an inner ring of mutable capsular ligament and an outer ring of muscle. Mechanical stimulation of a spine base results in the stiffening of its capsular ligament and the inclination of adjacent spines towards the stimulated spine; inclination of adjacent spines can occur only through the coordinated contraction of their muscles and destiffening of their ligaments.

5.3.3. Neuropeptides and Other Peptides

Extracts of the sea-cucumber Apostichopus japonicus contain myoactive neuropeptides that cause muscle relaxation or contraction and change the stiffness of sea-cucumber dermis: the amidated pentapeptide NGIWYamide causes stiffening and the cyclic peptide stichopin (DRQGWPACYDSKGNYKC) suppresses the stiffening effect of ACh on the dermis [196]. NGIWYamide- and stichopin-like immunoreactivity has been detected in nerve fibers located within the dermis [190,197]. Stichopin-like immunoreactivity was also detected in cells located in the dermis and in other collagenous tissue [197]. The vasopressin/oxytocin-type neuropeptide crinotocin causes nervously mediated destiffening of ligaments at the mobile diarthrial arm joints of the featherstar Antedon mediterranea but has no effect on the ligaments at the syzygial joints where autotomy occurs. Cells expressing crinotocin precursor transcripts are present in the main motor nerve of the arm and at the base of nerve branches that innervate the arm muscles and diarthrial ligaments [198]. The injection of another neuropeptide—ArSK/CCK1—into the starfish Asterias rubens promotes arm autotomy, which is likely to be due partly to its contractant effect on the tourniquet muscle that constricts the arm at the detachment plane during autotomy. However, as ArSK/CCK1-immunoreactive cells are present in the inter-ossicular collagenous tissue of the body wall, it may also be involved in initiating the destabilization of this tissue during autotomy [187,199].

The heptapeptides holokinin-1 and holokinin-2 (PLGYMFR and an oxidized derivative, respectively) cause destiffening of A. japonicus dermis. Analysis of A. japonicus transcriptome sequence data has revealed that the holokinins are not neuropeptides, as originally supposed [196], but fragments of the C-terminal region of a 5α type collagen [175]. As mentioned above (see Section 5.2), this may help to explain how holokinins influence the mechanical properties of MCT. However, the physiological relevance of holokinins in sea-cucumbers is unclear, since they could be a product of collagen degradation generated when body wall extracts are prepared.

5.3.4. Celomic Factors

The celomic fluid of intact and untreated echinoderms contains separate chemical factors that stiffen or destiffen isolated MCT preparations in vitro. These factors have been detected in representatives of all classes except the Crinoidea (no members of which have been investigated), and their effects appear not to be species- or class-specific [52,69,139]. The chemical nature, source, and physiological significance of these factors remain unknown. It has been conjectured that they are neurotransmitters that modulate the activity of JLCs [52].

The above factors have been reported only in sea-cucumber species from the orders Synallactida and Holothuriida. Sea-cucumbers from the order Dendrochirotida demonstrate defensive auto-evisceration of the retractable anterior portion of the body (the ‘introvert’) and the internal organs. This requires the destabilization of the introvert dermis and of tendons connecting the introvert retractor muscles to the longitudinal body wall muscles. Celomic fluid from eviscerating dendrochirotids contains a factor that induces evisceration when injected into intact animals and which destabilizes the introvert dermis and retractor tendons. This ‘evisceration factor’ is absent from the celomic fluid of intact animals and is a small molecule with a molecular mass of ca. 150 Da [200,201].

The celomic fluid of scalded starfish induces autotomy when injected into intact starfish, an effect that is not produced by celomic fluid from intact and untreated animals [202]. An ‘autotomy-promoting factor’ (APF) has been isolated from the celomic fluid of scalded starfish and partly characterized. It is likely to be a peptide with a molecular mass of 1100–1200 Da [203,204] and therefore possibly an ortholog of the autotomy-promoting neuropeptide ArSK/CCK1 (molecular mass 1613 Da) (see Section 5.3.3) [198]. Like that neuropeptide, starfish APF may activate MCT destabilization, since it causes the nervously mediated rupture of isolated preparations of the basal arm autotomy plane of Pycnopodia helianthoides [54]. As these investigations have involved only starfish species belonging to the family Asteriidae, the wider significance of APF is unclear.

5.4. Force Generation

It has been demonstrated that certain mutable collagenous structures in sea-urchins, sea-lilies and featherstars can generate contractile force. As well as increasing the tensile strength and stiffness of the capsular ligament of the sea-urchin spine joint (see Section 5.3.2), ACh also causes isolated preparations of it to shorten and develop tensile force [61,205]. Both the contraction and relaxation phases of the contractile response are very fast, the contraction phase reaching up to 90% of the maximum value in under 1 s (Figure 9A) [206]. Force generation in this case depends on myocytes that are distributed between the bundles of collagen fibrils and constitute 1–3% of the total cross-sectional area of the ligament. The functional significance of the myocytes and their contractile activity has not been established. It has been demonstrated experimentally that they are not responsible for varying the passive stiffness of the capsular ligament [207]. They may help to reshorten extended ligament fibers or they may operate synergistically with the spine muscle, for example during re-erection of the spine.

Though lacking myocytes, mutable ligaments in the cirri and arms of sea-lilies and featherstars have the capacity for active contractility. Cirri are claw-like appendages that have an endoskeleton consisting of a single series of interarticulating ossicles connected by ligaments. They are attached to the aboral surface of featherstars and the stalk of some sea-lilies. Those connected to the stalk of the sea-lily Metacrinus rotundus bend upwards when the stalk or the cirri themselves are stimulated mechanically. Cholinergic agonists induce slow force generation in isolated cirri undergoing stress relaxation (Figure 9B). This is not due to the passive elastic recoil of a structural component that was previously in an extended state, since cirral preparations can develop force even if they show partial stress relaxation before they are stimulated [133,209,210].

The bending of sea-lily and featherstar arms occurs at mobile hinge joints. Within each joint, there is a fulcral ridge, below which is a single ligament and above which are paired ligaments and paired muscles. Although contraction of the muscles bends the arm upwards, the power stroke that brings about locomotion by swimming, crawling, or climbing depends on the downward flexion of the arm and must therefore be accomplished by the ligaments. The aboral (sub-fulcral) ligament of featherstars has been shown to shorten slowly and generate a force of up to ca. 5.6 kPa [210,211]. By measuring stiffness and shortening simultaneously in the aboral ligament of a sea-lily, Motokawa et al. [208] demonstrated the independence of its passive mechanical properties and contractile activity (Figure 9C–F). However, as contracting ligaments are usually in the destiffened state, their passive and active mechanical behaviors appear to be coordinated.

No information is available on the mechanism of active force generation in featherstar and sea-lily ligaments. Birenheide & Motokawa [210] speculated that this could be related to the contraction of hydrogels that results from changes in their fluid phase [212].

6. Biological Significance of MCT

6.1. Energy-Sparing Postural Fixation

Since echinoderm collagenous tissue was first shown to undergo reversible stiffening and destiffening, it has been conjectured that this capacity is used to maintain the posture of the body or its appendages for extended periods of time as an alternative to tonic muscular contraction, and that MCT thus serves an energy-sparing function [24,25,138,163]. Prolonged postural fixation is employed primarily in feeding mechanisms and as a defensive strategy. Sea-lilies, featherstars, basket stars, brittlestars and starfish belonging to the family Brisingidae are suspension feeders with long arms, which, when collecting food items, are extended in fixed postures often for hours. In addition, during such feeding activity, sea-lilies and featherstars are anchored to their substrate by articulated appendages: a stalk or claw-like cirri, respectively [213,214,215,216,217]. Examples of defensive postural fixation are provided by sea-cucumbers in the family Stichopodidae that survive months of dormancy sheltering under stones while in a contracted and stiffened state [218,219]; sea-urchins that wedge themselves into holes and crevices in rock by means of their immobilized spines [73,220]; and brittlestars whose whole body ‘freezes’ and becomes rigid when they are severely disturbed [221].

There can be no doubt that MCT alone is responsible for the postural fixation of those structures that lack muscles (apart from myocytes associated with epithelia), i.e., sea-lily stalks, featherstar cirri and stichopodid sea-cucumber dermis [107,222,223]. However, although mutable ligaments are present at sea-urchin spine joints and in the arms of sea-lilies, featherstars, brittlestars, and starfish (Table 1), these anatomical structures also contain substantial muscles that actuate their movement and, theoretically, could also maintain posture by tonic contraction. This can be discounted in the case of sea-urchin spine joints, since the defensive ‘freeze response’, in which direct touch invokes prolonged immobilization of the spine in the upright position, persists after the spine muscle has been completely transected and thus must be achieved only through the stiffening of the capsular ligament [195]. Electrophysiological evidence also indicates that the rigidity of the arms of brittlestars undergoing defensive ‘freezing’ is unlikely to depend on continuous contraction of their intervertebral muscles. Intracellular recording has shown that, after initial excitatory post-synaptic activity in hyponeural motor neurons, the frozen posture is maintained in the absence of any apparent neural activity [221].

Protracted MCT stiffening also has a role in sea-urchin dental mechanics. Each sea-urchin tooth is attached to a ‘jaw’ by a tooth ligament. To compensate for wear caused by scraping hard surfaces, the tooth grows by the addition of new material at its dorsal end and, when the animal is not feeding, moves slowly in a ventral direction, which is permitted by the destiffened state of the tooth ligament. When the animal is feeding, the tooth is subjected to mechanical loading and, to prevent its dorsal displacement, the ligament switches to a stiffened state [64,124,134].

Investigations comparing the oxygen consumption rates of MCT and muscle have confirmed that fixing posture by the passive stiffening of MCT rather than muscle contraction is likely to confer considerable energetic advantages. As shown in

Table 7. Oxygen consumption rate (VO₂) of the dermis and longitudinal body wall muscle of the sea-cucumber Actinopyga mauritiana and of the dermis and longitudinal carinal muscle of the starfish Linckia laevigata. Dermis samples in the standard state and relaxed muscle preparations were in normal artificial seawater (ASW); dermis samples in the stiff state (see Section 5.1.2) and contracted muscle preparations were in ASW containing 100 mmol L⁻¹ K⁺ ions; in addition, a second batch of contracted sea-cucumber muscle preparations (asterisk) was in ASW containing 1 mmol L⁻¹ acetylcholine. Data from refs. [224,225].

Species	Average VO2 (µL·g−1·h−1) ± SD
	Dermis		Muscle
	Standard	Stiff	Relaxed	Contracted
Actinopyga mauritiana	1.61 ± 0.51	2.45 ± 1.34	9.21 ± 3.28	23.5 ± 5.8; 18.4 ± 3.6 *
Linckia laevigata	0.56 ± 0.12	0.78 ± 0.19	12.44 ± 5.54	29.30 ± 17.09

, the VO₂ of the contracted longitudinal body wall muscle of a sea-cucumber was found to be 7–10 times higher than that of its stiffened dermis, and the VO₂ of the contracted longitudinal carinal muscle of a starfish was nearly 40 times higher than that of its stiffened dermis. More precise comparisons of energy costs, taking into account the stress that contracted muscle and stiffened MCT can support (inferred from their mechanical stiffness), suggested that postural fixation of the body by the dermis would require only 1.4% of the energy needed if the same supporting force were provided by muscle alone in the sea-cucumber and 0.03% of the energy needed by muscle alone in the starfish [224,225].

6.2. Autotomy and Fission

The term ‘autotomy’ is used herein to denote the adaptive detachment of animal body parts where this serves a defensive (usually anti-predator) function, is achieved by an intrinsic breakage mechanism, and is neurally mediated [155,226]. The capacity for autotomy occurs widely throughout all five extant echinoderm classes and involves the disconnection of a diverse range of anatomical structures—from small appendages like featherstar cirri and sea-urchin pedicellariae, to complex body regions such as the arms of stellate echinoderms (including brittlestars, featherstars and starfish), and to potentially more than half of the body of apodid sea-cucumbers [86,155,182,227].

Echinoderm autotomy planes always transect mechanically significant collagenous components. In all investigated cases, which include examples from each echinoderm class, it has been found that these components consist of MCT that undergoes irreversible destabilization during autotomy (Table 1) [155]. Evidence for such a detachment mechanism has been obtained both from the direct observation of autotomizing structures [51,227,228,229] and by mechanical testing methods applied to isolated MCT preparations in vitro [53,59,70,76].

In all confirmed examples of echinoderm autotomy, anatomical structures are detached quickly (within less than a second to a few minutes) at a preformed, morphologically distinguishable autotomy plane at which there may be adaptations for minimizing tissue damage, sealing fluid compartments, assisting wound repair, and promoting regeneration [155]. There are, however, examples of echinoderm detachment phenomena occurring in response to external agents at sites that lack any autotomy-specific features but are associated with mutable collagenous structures normally involved in only reversible changes in stiffness. For example, although there are no reports of their spines undergoing rapid defensive self-detachment, brittlestars and sea-urchins can shed spines to which tags have been attached for experimental purposes, a process that can take hours to days and results from the drastic destiffening of their spine-joint ligaments [62,78]. Such occurrences have been described as ‘opportunistic self-detachment’ to distinguish them from autotomy sensu stricto [155].

Fission—asexual reproduction by division of the body—occurs in certain adult brittlestars, starfish, and sea-cucumbers [230] and in certain larval starfish [231]. Direct observation of splitting animals and experimental evidence obtained from isolated tissue preparations in vitro indicate that fission in adult echinoderms is likely to be assisted by the destiffening of MCT—predominantly body wall dermis—at the plane of separation [72,232,233]. This has also been confirmed by a comparative transcriptomic analysis of tissues from intact specimens of a fissiparous sea-cucumber and from specimens undergoing fission [93]. It is notable that the latter investigation [93] hints that proteases (other than MMPs) may have a greater role in fission-related dermal destabilization than has been proposed for reversible changes in dermal stiffness (see Section 5.2). There is morphological evidence that at least one form of asexual reproduction in starfish larvae (detachment of the anterior portion of the pre-oral lobe in bipinnariae of Luidia spp.) occurs at a preformed breakage plane [231]. However, although this process must involve the rupture of ECM structures in the larval body wall [234], it is not known if this is facilitated by a pre-weakening mechanism.

6.3. Sea-Cucumber Dermal Autolysis

Sea-cucumber dermal autolysis (SCDA), also referred to as ‘liquefaction’, ‘local degeneration’, and ‘melting’, is observed when certain sea-cucumbers are touched firmly or gripped. After an initial stiffening reaction, the stimulated area softens progressively over a period of minutes until it can be penetrated easily, allowing extrusion of the internal organs (evisceration) [25,235]. In contrast to such localized responses, persistent mechanical stimulation or exposure to UV radiation, high temperatures, or irritant chemicals can convert large areas of the body wall into a viscous flowing mass (Figure 10) [236,237,238]. SCDA is a serious cause of deterioration in the quality of commercially harvested edible sea-cucumbers, especially if they are subjected to prolonged transportation, insensitive handling, or unsuitable storage conditions [239].

SCDA has been regarded as a form of autotomy [25,236]. However, autotomy by definition involves anatomical detachment (see previous Section) and this is not an outcome of SCDA. In addition, some species that demonstrate SCDA—notably Stichopus chloronotus and S. horrens—also have the capacity to detach discrete areas of (non-autolyzed) body wall in response to predator attack. During this process the mesentery surrounding the internal viscera is not punctured and evisceration does not occur [240,241]. This is undoubtedly a defensive autotomy response and is clearly a separate process from SCDA.

Because of its economic implications, the biological pathways responsible for SCDA have been the subject of intense scrutiny, with most of the focus being on proteolytic mechanisms. The activity of several endogenous proteases increases in disintegrating dermis [242], though autolysis is blocked most effectively by MMP inhibitors [243] and only MMP activity accomplishes the complete disaggregation of collagen fibers into smaller fibril bundles and fibrils (together with the degradation of interfibrillar bridges), which appears to be the basis of the destabilization process [243,244,245,246,247]. As several MMP genes are also significantly upregulated in autolyzed dermis [248], it seems that MMPs are the key endogenous proteases responsible for SCDA [243].

The sources of the proteases involved in SCDA include dermal cells containing multivesicular bodies: one of the implicated enzymes—cathepsin L—is released from these ‘vacuole cells’ during autolysis in Apostichopus japonicus [249]. Though not reported for A. japonicus, ‘vacuole cells’ in a related species—Stichopus chloronotus—have processes that are distributed throughout the dermis, mostly isolated from other cellular components, though some occur in bundles with juxtaligamental and other neuronal processes to which they are closely adjacent [250]. These ultrastructural observations complement physiological evidence that SCDA is nervously mediated: pretreatment with anesthetic agents blocks autolysis in Isostichopus badionotus [251].

Three aspects of SCDA suggest it has affinities with ‘conventional’ mutability phenomena and is not a pathological process. (1) It results from targeted proteolysis that disaggregates both collagen fibers and their component fibrils but does not damage monomeric collagen [245,246]. This resonates with the model of Tamori et al. [178], which proposed that the reversible destiffening of holothurian dermis involves collagen fiber and fibril disaggregation (see Section 5.2). (2) MMPs contribute to SCDA and are hypothesized to have a role in the reversible destiffening of mutable collagenous structures in general [174]. (3) There is morphological and physiological evidence that SCDA, like the routine tensile changes undergone by MCT, is nervously mediated. It is therefore feasible that, as surmised by Motokawa [25], localized dermal autolysis evolved as an adaptive predation-avoidance tactic that makes it difficult for an attacker to achieve a firm grip, whereas the non-localized, potentially fatal expression of SCDA that affects wide areas of body wall is caused by the non-adaptive over-activation of the proteolytic softening mechanism.

6.4. Regeneration

Echinoderms are renowned for their regenerative abilities and have provided several productive models for regeneration research [252,253,254,255]. Although available evidence is sparse, it suggests that autotomy is the most common proximate cause of structural loss in echinoderms in their natural environment [155]. This implies that most echinoderm regeneration events are preceded by autotomy and occur at the retained side of a fractured autotomy plane. As a result of the presence of MCT at all echinoderm autotomy planes, the post-autotomy wound surface invariably includes disrupted MCT. There is evidence that the JLCs in disrupted MCT can have a role in regeneration.

The visceral mass of featherstars is connected to the endoskeleton by mesentery-like septa. It is highly likely that collagenous tissue in the septa has mutable properties that facilitate defensive detachment of the visceral mass [60]. It has been reported that, after loss of the visceral mass in Himerometra robustipinna and Dichrometra (previously Lamprometra) palmata, JLCs migrate from the ruptured septa to the wound surface and start a process of direct transdifferentiation (there being no intermediate stage of dedifferentiation). They undergo a series of ultrastructural changes that, after 24 h, transform them into enterocyte (gut epithelial cell) precursors with developing microvilli on their apical surface and secretory granules characteristic of mature enterocytes in their apical region [256,257]. In D. palmata, JLCs also transform into epidermal cells in the regenerating body wall of the visceral mass [257]. The presence of a reservoir of cells with such transformation potential close to the regeneration site may account for the rapid restoration of a functioning visceral mass, which can be accomplished in as little as 4–7 days after autotomy [256,257]. Furthermore, a comparative investigation of visceral mass regeneration in a taxonomically diverse range of 15 featherstar species identified a positive correlation between JLC abundance and regeneration rate [258].

The cell plasticity implied by these observations is plausible, since mammalian neurons can transform into epithelial cells in vitro, though this involves a dedifferentiation stage [259]. However, JLCs do not contribute to visceral mass regeneration in all featherstars, since the enterocyte precursors in Antedon mediterranea are derived from celothelial cells that dedifferentiate then redifferentiate [260]. In addition, there are no reports of JLCs being involved in the regeneration of any other echinoderm structures. On the other hand, the cellular mechanisms of only a very small proportion of all echinoderm regeneration processes have been studied in detail, and many of these relate to events following surgical amputation rather than induced autotomy, a procedure that damages the autotomy plane or avoids it altogether and is known to result in anomalous regeneration [261,262].

6.5. Pathology

6.5.1. Sea Star Wasting Disease

The condition known as sea star wasting disease (or syndrome) (SSWD) is characterized by a set of grossly abnormal signs that include body deflation or loss of turgor, arm loss, body wall lesions, and abnormal behavior (Figure 11A,B). It has been reported for over 100 years, affects dozens of starfish species, and may be common, but, despite over two decades of intensive research, distinct causative agents have not been identified [263,264].

Whatever its etiology, the pathogenesis of SSWD involves the starfish body wall, including its mutable dermis. One of the earliest signs of SSWD is ‘inflammation’ of the dermal MCT, which refers to its infiltration by hemocytes and its edematous appearance resulting from increased spacing between its constituent collagen fibers (Figure 11C,D). These changes precede epidermal ulceration and the softening and eventual disintegration (‘melting’) of the body wall [266,267,268]. Gene expression studies comparing symptomatic and asymptomatic animals have revealed differences in the expression of collagen, cell adhesion, and tissue remodeling genes (including increased expression of MMPs), suggesting enhancement of degradative processes in the MCT of diseased individuals [269,270]. These investigations have also provided evidence that the nervous system is affected. For example, in Pycnopodia helianthoides, amongst other changes, reduced expression of acetylcholinesterase and a muscarinic receptor indicate that ACh signaling pathways may be disrupted [271]. It has also been surmised that the upregulation of RNA processing genes detected in symptomatic Pisaster ochraceus could be a sign of neurological dysfunction [270]. On the basis of such findings, it has been repeatedly conjectured that the body wall changes observed in SSWD result from disturbance of neural pathways controlling MCT tensility [263,267,268,270,271]. This is highly speculative and leaves unexplained the pathogenetic connection between putative neurological dysfunction and the inflammatory and degradative processes occurring in the initial stages of SSWD.

6.5.2. Sea-Cucumber Skin Ulceration Disease

Skin ulceration disease (or syndrome) (SKUD) was first reported in 1998 and has been observed worldwide mainly in farmed edible sea-cucumbers, causing up to 100% mortality in cultured stocks [272,273]. It is therefore, like sea-cucumber dermal autolysis (see Section 6.3), of great commercial importance.

The symptomatology of SKUD resembles that of SSWD in that the integument of affected animals develops white lesions where destruction of the cuticle, epidermis, and outer dermis has exposed the deeper dermis (Figure 12). The lesions widen and the dermis undergoes extreme softening prior to being completely punctured [273,274,275]. Further similarities to SSWD have been revealed by a histopathological investigation of Isostichopus badionotus, which demonstrated that changes in the dermis of lesions—disorganization of its collagen fiber microarchitecture, ‘liquefactive necrosis’, and ‘anhistic’ areas (possibly indicating edema)—can occur before the epidermis is damaged [276].

Whilst there is strong evidence that some cases of SKUD are caused by highly infectious bacterial or viral pathogens, the initial causes of others are uncertain and may be abiotic, with opportunistic pathogens infecting lesions secondarily [273,275]. A direct role for MCT in bacterial pathogenesis has been suggested by a proteomic analysis of the body wall of Apostichopus japonicus specimens in which SKUD was induced experimentally by exposure to Vibrio splendidus bacteria. This showed that 5α collagen was one of only two proteins that were downregulated at all stages of disease progression [277]. Since tissue colonization and the establishment of infections in mammals can involve interactions between bacterial pathogens and ECM components [278,279], it has been argued that 5α collagen might be a target protein of causative pathogens, its downregulation representing a protective response [277]. However, if 5α collagen is involved in interfibrillar crosslinking (see Section 5.2), its downregulation could also be a factor in the ECM disorganization that contributes to dermal lesion formation.

7. Applications

Interest in the potential biomedical and other applications of MCT has been steadily growing since Trotter et al. [169] first introduced the concept of MCT-inspired synthetic systems. Two types of applications have emerged: biomimetic applications, which exploit information on MCT organization and the mechanisms underpinning its mechanical adaptability but do not incorporate any MCT components; and biotechnological applications, which use extracted MCT components for deployment in synthetic materials [17].

7.1. Biomimetic Applications Inspired by MCT

The system envisaged by Trotter et al. [169] was a ‘hybrid’ biomaterial comprising fibrils extracted from holothurian dermis and a synthetic interfibrillar matrix. Capadona et al. [280] were the first to assemble and test a wholly synthetic mechanically adaptive material inspired by sea-cucumber dermis, which could be used in the construction of implanted intracortical microelectrodes. The model employed tunicate (sea-squirt)-derived cellulose nanofibers immersed in a polymeric matrix (ethylene oxide-epichlorohydrin 1:1 polymer or polyvinyl acetate). Cellulose nanofibers tend to aggregate via hydrogen bonding due to the high density of strongly interacting surface hydroxyl groups [281]. This interaction can be switched off by water through competitive hydrogen bonding that reduces hydrogen bonding between the nanofibers and destiffens the composite reversibly. Restiffening occurs when water is removed (Figure 13A,B) [280,282]. In contrast to the reversible stiffening/destiffening system of this first MCT-inspired synthetic composite, irreversible stiffening (which is shown by sea-cucumber Cuvierian tubules: see Section 5.1.2) is the key property of one of the most recent—a calcium phosphate-based cement designed for bone repair, which incorporates cellulose nanofibers and polyvinyl alcohol. This can be injected as a hydrated slurry in which water prevents hydrogen bonding between the nanofibers and the polyvinyl alcohol. When the slurry solidifies, the consequent reduction in the water phase permits hydrogen bonding between the additives, resulting in a material with high strength and toughness [283].

The sea-cucumber dermis has served repeatedly as a conceptual model for the development of hydrogels, which are three-dimensional crosslinked polymeric networks swollen with water. Hydrogels can respond to external and internal physical or chemical stimuli, such as temperature, electric fields, chemical compounds, and pH (for review, see [285]). They have been employed as injectable biomaterials, tunable surfaces for cell sheet engineering, sensors, and actuators (see, e.g., [286]). As indicated by in vitro cytotoxicity tests and preliminary subcutaneous implantation, supramolecular polymer hydrogels can be both biocompatible and autolytic in vivo, and could be used for such purposes as intestinal drug delivery or injectable filling to facilitate the suturing of small vessels.

Ideas derived from sea-cucumber dermis mutability have also assisted the development of shape-memory polymer-cellulose nanofiber composites that can adopt one or more temporary shapes while retaining the ability to return to their original shape [282,287]. Early models employed water movement to regulate hydrogen-bonding interactions in the nanocomposite framework, while later models responded to environmental stimuli, such as temperature or pH changes [288,289,290].

At the most general level, the fast stiffness modulation of MCT, achieved by adjusting interfibrillar stress transfer, has also inspired models of stiffness tuning that involve completely unrelated molecular mechanisms, such as reversible thermoplastic softening, as demonstrated by polymers for coating heating elements [291] and the thermoplastic mesh of a shape-changing material that mimics starfish dermis [50].

MCT has been a source of design concepts for soft actuators with uses in diverse fields including flexible electronics, soft robotics, and biomedicine [212]. Of particular interest are water-driven self-operating actuators that perform movements by means of a swelling/deswelling process. Choi et al. [292] developed a dynamic poly(N-isopropylacrylamide) hydrogel whose elastic modulus could be altered via hydration/dehydration and which had shape-memory properties. This was incorporated into a programmable and ultrafast water-driven self-operating soft actuator (SITuator), which was based on hydrogel stiffness changes obtained by varying crosslinker concentration.

Other synthetic analogs of sea-cucumber dermis have been trialed for employment in soft robotic applications. One such example [293] is an impact-protective supramolecular polymeric material (SPM) characterized by impact-hardening and reversible softness-stiffness switchability. This SPM consists of chains of a triblock copolymer (poly(ethylene glycol)-block-poly(propylene glycol)-block-poly(ethylene glycol) to which 2-ureido-4-pyrimidone (UPy) dimers are attached that are stabilized by quadruple hydrogen bonding. The UPy dimers form crosslinked domains that, when subjected to a sudden impact, undergo reversible stiffening, thereby reducing deformation, possibly through the rapid reconstitution of hydrogen bonds after initial impact-induced hydrogen bond disruption.

7.2. Biotechnological Applications Employing MCT Components

Collagen obtained from animal sources is used widely in the production of biomedical materials for the management of human diseases and defects, particularly where tissue repair and regeneration need to be promoted. Extracted collagen can be assembled to form 2D membranes and 3D meshes, which are subjected to crosslinking procedures to increase their mechanical stability and resistance to enzymatic degradation. Whilst most of the collagen is currently mammalian in origin, there is increasing interest in replacing this with collagen from marine invertebrates, including echinoderms, for various reasons, such as marine collagen’s relative cheapness, ready availability in large quantities, immunological advantages, and low risk of disease transmission [43,294,295]. Since it is very difficult to isolate intact fibrils from the collagenous tissue of postfetal vertebrates (see Section 5.2), most mammalian collagen is used in its hydrolyzed (acid-solubilized) form, which compromises its mechanical performance. Echinoderm collagen, however, can be extracted as intact fibrils using mild, non-denaturing methods that leave in place D-periodically distributed surface GAGs, which may help to stabilize synthetic assemblages of the fibrils [284,295,296].

The main source of echinoderm collagen for prospective biotechnological investigations has been sea-urchin peristomial membranes (PMs), which consist largely of a mutable dermal layer [66,84] and are available in abundance as a waste-product left by the harvesting of sea-urchin gonads for human consumption. Both 2D membranes and 3D scaffolds derived from PM collagen have been investigated as potential alternatives to commercially available bovine collagen-based biomaterials (Figure 13C,D). PM-derived sheets have been shown to have good biocompatibility in terms of enabling the growth of mammalian mesenchymal stromal cells and dermal fibroblasts and could be used for cell culture and human health applications such as guided tissue regeneration [295,296]. Thin films of fibrils extracted from another MCT—sea-cucumber dermis—have reversible stimulus-responsive behavior. At 25 °C the films are translucent and soft; when subjected to an increasing temperature, they become opaque and brittle at 100 °C, reverting to their original state when the temperature drops back to 25 °C [297].

PM-derived collagen is also being used in the development of a new bilayered skin substitute intended as an alternative management option to skin grafting for wound repair. This consists of a membrane that functionally mimics epidermis by reducing water loss and preventing bacterial infiltration, and a scaffold that functions as a mechanically stable dermis. This ‘dermal template’ has been shown to be an effective substrate for fibroblast infiltration, survival, and proliferation [298] and to promote experimental wound healing in sheep and rats, resulting in outcomes similar to those achieved by commercially available artificial skin that incorporates bovine collagen [299,300].

A range of decellularized bioscaffolds that retain their extracellular structural framework and are derived from human and other mammalian tissues is now available and widely applied in tissue engineering and regenerative medicine [301]. Regarding MCT, Goh and Holmes [302] provided an exhaustive theoretical framework pertaining specifically to the use of sea-urchin structures for tissue engineering purposes. However, this application area has been explored so far only in pilot studies employing decellularized dermis from sea-urchin PMs. These PM-derived scaffolds have been investigated with the goal of advancing the development of echinoderm cell cultures, a research field that is still largely unexplored. It is possible that they could eventually have a role in human therapeutics [303].

8. Gaps in Knowledge

The nervously mediated mechanical adaptability of echinoderm collagenous tissue is one of the most unusual phenomena yet encountered in the field of animal biology. However, despite 50 years of research effort, the understanding of it is far from complete and many aspects remain unexamined or poorly characterized. This applies both to the extracellular components of MCT and its key cellular components—the juxtaligamental cells.

8.1. Extracellular Components and Mechanisms of Tensile Change

Significant progress in elucidating the molecular basis of the unique mechanical behavior of MCT—and in realizing its full potential for biomedical and technological exploitation—is unlikely to be made unless there is an exhaustive characterization of its molecular and supramolecular organization, particularly with regard to those components responsible for interfibrillar cohesion and stress transfer. This has not been achieved so far for any mutable collagenous structure. Whilst several possible constitutive or regulatory factors (e.g., stiparin and tensilin, respectively) have been isolated from sea-cucumber dermis, their exact contributions to interfibrillar crosslinking and its dynamics have yet to be ascertained. To put this in context, however, it should be remembered that there is still uncertainty as to which molecular components transfer interfibrillar shear forces in mammalian tendons [157].

Concerning mechanisms of passive tensile change, the three-state model of Motokawa and Tsuchi [135] remains an inspirational paradigm that gave rise to the specific stiffening hypothesis of Tamori et al. [178]. It needs to be emphasized that the three-state model was derived from the mechanical behavior of sea-cucumber dermis and that most of the endogenous factors that influence MCT properties have been isolated from this tissue alone. The relevance of the three-state model for other mutable collagenous structures that show reversible changes is therefore unclear. It is certainly not applicable to those structures that show only irreversible stiffening or destabilization, least of all to the non-fibrillar basement membrane-derived autotomy tendons of brittlestars. It has also become apparent that sea-cucumber dermis cannot be regarded as a representative fibrillar MCT. This can be inferred from the unique presence of fucosylated chondroitin sulfate [49] and of morula cells [110,128], and the unusual features of its juxtaligamental supply: separation from the ECM by a basement membrane, lack of JLC aggregations, and ectoneural innervation [46]. The discovery that sea-cucumbers possess many more TIMP genes than do the other four echinoderm classes [304] may be a further indication of the distinctiveness of their collagenous tissue. These considerations underscore the need for a radical shift in emphasis to other mutable collagenous structures that are amenable to both comparative ‘-omics’ approaches and in vitro functional investigations. Potential experimental models include the body walls of asteriid starfish [86], brittlestar ventral arm plate ligaments [76], and featherstar cirral ligaments [56].

The ability of some featherstar and sea-lily ligaments to develop contractile force adds another layer of complexity to the MCT phenomenon. However, this aspect appears to have been totally neglected for 20 years [208] and the mechanism of force generation remains unknown. Information derived from biomimetic analogs, i.e., contractile hydrogels [292], together with the evidence from different mutable collagenous structures that alterations in passive tensile properties can be accompanied by changes in ECM water content and fibril packing density [98,176,177,178], could be the foundation of testable hypotheses.

8.2. Juxtaligamental Cells

The ontogenesis of JLCs has not been investigated. Elucidation of their developmental origins in different classes would clarify their relationship with the nervous system and help to determine if they are homologous across the phylum [46]. Several experimental models are potentially suitable for this purpose, including the normal (i.e., not regenerating) growing tips of starfish, featherstar and brittlestar arms, the early post-metamorphic stages of sea-urchins [305] and sea-cucumbers [306], and the intrabursal juveniles of viviparous brittlestars such as Amphipholis squamata [307]. Concerning the morphogenesis of juxtaligamental systems, nothing is known of the processes by which mutable collagenous structures come to be populated with JLCs. ECM components are known to regulate the proliferation, differentiation, and migration direction of vertebrate and invertebrate neurons [308,309], and echinoderms possess many homologs of the participating neural adhesion proteins [95]. Insight into this aspect could be obtained by comparative gene expression analyses, taking advantage of experimental models in which a mutable collagenous structure (that contains JLCs) is adjacent to a non-mutable collagenous structure (that lacks JLCs). Accessible and tractable examples of such models are the capsular and central spine ligaments of cidaroid sea-urchins [81] and the autotomy and non-autotomy tendons of brittlestar intervertebral muscles [75].

Whilst circumstantial evidence strongly implicates JLCs as the effectors responsible for the rapid modulation of MCT tensility, a crucial piece of evidence is missing: the demonstration of a correlation between externally provoked, nervously mediated tensile change and the release from JLCs of a previously stored chemical factor that produces the same tensile change when isolated preparations are treated with it in vitro. This is symptomatic of the very limited information that is available on JLC activities and functions, current concepts of which have been derived almost exclusively from ultrastructural observations. This deficiency could be partly redressed by means of transcriptomic and/or metabolomic profiling, an important aim of such an approach being to define JLC secretomes [310,311]. A recent transcriptome-wide gene expression investigation of the brittlestar Ophiocoma wendti is a first step in this direction. A comparison of this animal’s inner arm core (which is enriched in JLCs) with its whole arm (containing a basal level of JLCs) and its stomach (lacking JLCs) led to the identification of 16 genes whose expression may be specific to JLCs [47]. It is feasible that this methodology could be applied to isolated JLCs because juxtaligamental nodes in the arms of some brittlestar species are surgically accessible [46,106].